0
$\begingroup$

I'm getting a TDR trigger after 4ms with a GL4.3 compute shader. tdrDelay is currently set to 30 secs, this obviously isn't long enough to get a timeout with the default Windows 11 tdrDelay value let alone my registery value, however the error message is code 7 (RTX A2000, drivers 528.24) which is the generic nVidia TDR occurance.

If there was a bad loop (infiniting) or something, I appreciate this could trigger a timeout, however can TDR's be triggered if the shader crashes (bad data, out of bounds etc) well before the tdrDelay is reached?

In the past I got different error messages if the shader crashes (this was on K2200D iirc, much old drivers, 3XX.XX), however the only error message I seem to get these days is a generic TDR error message (latest nVidia drivers)?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

Yes, turns out an index out of range index was the culprit when attempting writing to a TBO in the last batch of a workgroup invocation.

I long for the day when compute shaders have better dedicated tools for debugging and drivers supply more descriptive runtime error messages.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.